LabelFormer: Object Trajectory Refinement for Offboard Perception from LiDAR Point Clouds

TL;DR

LabelFormer is a novel, efficient trajectory refinement model that leverages self-attention for improved auto-labeling of LiDAR point clouds, enhancing perception accuracy for self-driving systems.

Contribution

It introduces a simple, temporal self-attention-based refinement approach that outperforms existing complex models in LiDAR auto-labeling tasks.

Findings

Outperforms existing refinement models significantly

Improves downstream detection performance when used for auto-labeling

Effective on both urban and highway datasets

Abstract

A major bottleneck to scaling-up training of self-driving perception systems are the human annotations required for supervision. A promising alternative is to leverage "auto-labelling" offboard perception models that are trained to automatically generate annotations from raw LiDAR point clouds at a fraction of the cost. Auto-labels are most commonly generated via a two-stage approach -- first objects are detected and tracked over time, and then each object trajectory is passed to a learned refinement model to improve accuracy. Since existing refinement models are overly complex and lack advanced temporal reasoning capabilities, in this work we propose LabelFormer, a simple, efficient, and effective trajectory-level refinement approach. Our approach first encodes each frame's observations separately, then exploits self-attention to reason about the trajectory with full temporal context, and finally decodes the refined object size and per-frame poses. Evaluation on both urban and highway datasets demonstrates that LabelFormer outperforms existing works by a large margin. Finally, we show that training on a dataset augmented with auto-labels generated by our method leads to improved downstream detection performance compared to existing methods. Please visit the project website for details https://waabi.ai/labelformer